07/10/2022
An article published on Academic describes a disease-specific research model, and the results of this study may be useful for disease-specific research and clinical staff.
The work involved in this article establishes the construction and multimodal reasoning process of a disease-specific knowledge graph (SDKGs).
Constructed SDKG-11, a SDKG collection of 11 categories including 5 cancers, 6 non-cancer diseases, 1 combined cancer, and 1 combined disease.
06/26/2022
Recently, evidence has shown that methylation patterns of ctDNAs help statistical models to diagnose with more reasonable accuracy than mutational patterns.
Despite these promising facts, early detection of cancers with methylation patterns of ctDNAs remains statistically difficult because the ratio of ctDNAs to cfDNAs is negligible.
Therefore, it is difficult to set a difference threshold on ctDNA ratios to successfully classify early-stage cancer patients. Which means, a high-throughput classifier was developed.
06/12/2022
scan is a type of X-ray scan that often helps diagnose fractures, heart disease, and emphysema. While CT scans can aid in disease diagnosis, and medical radiation is the second largest source of radiation.
The Indian laboratory said that this is the first time to use Transformer for medical image denoising, and it can benefit from the patch embedding operation of Vision Transformer:
(1) Their paper proposes a new edge-enhanced Transformer model ( ) for medical image denoising, which combines a learnable Sobel filter for edge enhancement, thereby improving the performance of the overall architecture.
(2) Demonstrates the effectiveness of residual learning in denoising. The experimental results show that residual learning is significantly better than traditional learning methods.
05/29/2022
case study assessing 's association with autism clinical phenotype:
A team from Harvard evaluate the autism-related disease node correlations in PrimeKG in two steps.
First, by performing entity resolution on autism concepts in all relevant raw data sources.
Second, by examining how these autism concepts correlate with autism Relationship between clinical subtypes.
